Add functionality for Pareto optimization

[thomaswmorris]

Pareto optimization is an alternative to scalarization where we explore the Pareto front (https://en.wikipedia.org/wiki/Pareto_efficiency).

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[mrakitin]

Looks good overall. I posted a few comments and suggestions below.

I think what is missing is an example (Jupyter/Sphinx) to demonstrate the Pareto front optimization. Can it be added, please?

[mrakitin]

What's better about this color map?

[thomaswmorris]

The values near 1 are hard to see on a white background. Turbo goes from red to blue and isn't white in the middle.

[mrakitin]

Can we add a comment/note about it next to the color map assignment?

[mrakitin]

Need an example. The "approving" review was preliminary.

[mrakitin]

Looks good, suggested a couple of minor corrections below, it's good to go after that.

Just to confirm - the learning process will still be the same, i.e. no special input is needed to declare we want to use the Pareto-efficient optimization. The plotting will help to understand the Pareto front and then one can pick some points from the front. Is my understanding correct?

[thomaswmorris]

Looks good, suggested a couple of minor corrections below, it's good to go after that.

Just to confirm - the learning process will still be the same, i.e. no special input is needed to declare we want to use the Pareto-efficient optimization. The plotting will help to understand the Pareto front and then one can pick some points from the front. Is my understanding correct?

Yup, nothing is changed.